It looks like Adtran and Digium are screaming the same thing. I'm going to
post the response I sent to Digium:

According to http://tools.ietf.org/html/rfc5589#page-26 (see the refer-to
on page 27) this is a valid transfer scenario. I have attached a valid
capture using FreeSWITCH as a gateway, and when the particular REFER that
is giving the Adtran GW problems comes up, it simply performs the refer and
marks the replaces in the INVITE (see highlights):

REFER sip:[email protected]:5080;transport=udp;gw=
sip.corp.ezuce.com SIP/2.0
From: <sip:[email protected]>;tag=4Ut224
To: "4092330016"<sip:[email protected]>;tag=Se415mQNX7D7D
Call-Id: 4d339693-a499-1230-729f-fad9dc40d221
Cseq: 3 REFER
Contact: <sip:[email protected]:5120;transport=tcp;x-sipX-nonat>
Referred-By: <sip:[email protected]>
Refer-To: <*sip:[email protected]?**
REPLACES=a0511239-ab6decb2-99d54f8f%40172.16.1.51%3Bto-tag%3DCA50B7BD-573B246%3Bfrom-tag%3DLkTYH8&REQUIRE=replaces&X-sipX-Authidentity=%3Csip%3A%7E%7Eid%7Epark%
40sip.corp.ezuce.com
%3Bsignature%3D509C36D7%253A%253Ad1b624445c9e26173fe0de17f291d62d%3E&REFERENCES=4d339693-a499-1230-729f-fad9dc40d221%3Brel%3Drefer
*>
References: [email protected];rel=xfer
Date: Thu, 08 Nov 2012 22:48:55 GMT
Max-Forwards: 19
User-Agent: sipXecs/4.6.0 sipXecs/park (Linux)
Accept-Language: en
Allow: INVITE, ACK, CANCEL, BYE, REFER, OPTIONS, NOTIFY, SUBSCRIBE
Supported: replaces
Proxy-Authorization: Digest username="~~id~park", realm="sip.corp.ezuce.com",
nonce="5c81f6a19cf6bd3bcaefdfe726e2db2e509c36d7",
uri="sip:[email protected]:5080;transport=udp;gw=
sip.corp.ezuce.com", response="a8fdc3b3eeb0f5e52ef91342aba9df3f",
cnonce="jbJLTS", qop=auth, nc=00000001
Via: SIP/2.0/UDP 172.16.1.5;branch=z9hG4bK-XX-0b96Qj3NNqiD6usxS_3GIL_emg
Via: SIP/2.0/TCP 172.16.1.5:5120
;branch=z9hG4bK-XX-003aW8u6STjzakMVEqA5D65UPg;id=18008-29
Content-Length: 0

SIP/2.0 202 Accepted
Via: SIP/2.0/UDP 172.16.1.5;branch=z9hG4bK-XX-0b96Qj3NNqiD6usxS_3GIL_emg
Via: SIP/2.0/TCP 172.16.1.5:5120
;branch=z9hG4bK-XX-003aW8u6STjzakMVEqA5D65UPg;id=18008-29
From: <sip:[email protected]>;tag=4Ut224
To: "4092330016" <sip:[email protected]>;tag=Se415mQNX7D7D
Call-ID: 4d339693-a499-1230-729f-fad9dc40d221
CSeq: 3 REFER
Contact: <sip:[email protected]:5080;transport=udp;gw=
sip.corp.ezuce.com>
Expires: 60
User-Agent: FreeSWITCH-mod_sofia/1.2.3
Allow: INVITE, ACK, BYE, CANCEL, OPTIONS, MESSAGE, INFO, UPDATE, REGISTER,
REFER, NOTIFY
Supported: timer, precondition, path, replaces
Allow-Events: talk, hold, conference, refer
Content-Length: 0

INVITE *sip:[email protected]* SIP/2.0
Via: SIP/2.0/UDP 172.16.1.90:5080;rport;branch=z9hG4bK2pQyFrXy5e8mN
Max-Forwards: 70
From: "4092330016" <sip:[email protected]>;tag=tQXt7F8rtg4SS
To: <sip:[email protected]>
Call-ID: 50251028-a499-1230-729f-fad9dc40d221
CSeq: 35871402 INVITE
Contact: <sip:[email protected]:5080>
User-Agent: FreeSWITCH-mod_sofia/1.2.3
Allow: INVITE, ACK, BYE, CANCEL, OPTIONS, MESSAGE, INFO, UPDATE, REGISTER,
REFER, NOTIFY
*Require: replaces*
Supported: timer, precondition, path, replaces
Allow-Events: talk, hold, conference, refer
*Replaces: [email protected]
;to-tag=CA50B7BD-573B246;from-tag=LkTYH8*
Content-Type: application/sdp
Content-Disposition: session
Content-Length: 255
X-FS-Support: update_display,send_info
*X-sipX-Authidentity: <sip:[email protected]
;signature=509C36D7%3A%3Ad1b624445c9e26173fe0de17f291d62d>
REFERENCES: 4d339693-a499-1230-729f-fad9dc40d221;rel=refer
*Remote-Party-ID: "4092330016" <sip:[email protected]
>;party=calling;screen=yes;privacy=off

Obviously the Adtran will not have knowledge of it, it's a call dialog on a
completely different SIP client. Refer-To does not use existing dialogs to
perform REFER. This is allowed per RFC 3515:

2.4.3 Accessing the Referred-to Resource

   The resource identified by the Refer-To URI is contacted using the
   normal mechanisms for that URI type.  For example, if the URI is a
   SIP URI indicating INVITE (using a method=INVITE URI parameter for
   example), the UA would issue a new INVITE using all of the normal
   rules for sending an INVITE defined in [1].




On Wed, Dec 12, 2012 at 11:12 AM, Bryan Anderson <[email protected]>wrote:

> So as I said before I am not well versed in the SIP protocol but this was
> Adtrans response.
>
> Thanks Bryan,
>
> I do not believe this to be an ADTRAN issue.  The call leg that is being
> replaced only exists on that PBX.  Check out the Refer message sent to the
> ADTRAN:
>
>  Rx: UDP src=10.0.31.5:5060 dst=10.0.31.4:5060
> 18:07:20.905 SIP.STACK MSG         REFER sip:10.0.31.4:5060;transport=
> UDP SIP/2.0
> 18:07:20.905 SIP.STACK MSG         From: <sip:[email protected]
> >;tag=NkynMZ
> 18:07:20.905 SIP.STACK MSG         To:
> <sip:10.0.31.4>;tag=3bd6630-7f000001-13c4-2e2d5b-dce8ab25-2e2d5b
> 18:07:20.906 SIP.STACK MSG         Call-Id:
> [email protected]
> 18:07:20.906 SIP.STACK MSG         Cseq: 3 REFER
> 18:07:20.906 SIP.STACK MSG         Contact: <sip:[email protected]:5120
> ;transport=tcp;x-sipX-nonat>
> 18:07:20.906 SIP.STACK MSG         Referred-By: <sip:[email protected]
> >
> 18:07:20.906 SIP.STACK MSG         Refer-To: <
> sip:[email protected]?REPLACES=a5320d9d-1d788972-8a9aa6d3%4010.0.27.106%3Bto-tag%3D1C951DC1-EB497A86%3Bfrom-tag%3Drp_5OC&REQUIRE=replaces&X-sipX-Authidentity=%3Csip%3A%7E%7Eid%7Epark%40inwsip.lcsnw.org%3Bsignature%3D50C7E6D8%253A%253A05936d0a35fdf946cde62d9a14e3e446%3E&REFERENCES=3bffb58-7f000001-13c4-2e2d5b-cf3f8dba-2e2d5b%4010.0.31.4%3Brel%3Drefer
> >
> 18:07:20.906 SIP.STACK MSG         References:
> [email protected];rel=xfer
> 18:07:20.907 SIP.STACK MSG         Date: Wed, 12 Dec 2012 02:07:20 GMT
> 18:07:20.907 SIP.STACK MSG         Max-Forwards: 19
> 18:07:20.907 SIP.STACK MSG         User-Agent: sipXecs/4.4.0 sipXecs/park
> (Linux)
> 18:07:20.907 SIP.STACK MSG         Accept-Language: en
> 18:07:20.907 SIP.STACK MSG         Allow: INVITE, ACK, CANCEL, BYE, REFER,
> OPTIONS, NOTIFY, SUBSCRIBE
> 18:07:20.907 SIP.STACK MSG         Supported: replaces
>
> Below is the Refer To from that message:
>
> Refer-To: <
> sip:[email protected]?REPLACES=a5320d9d-1d788972-8a9aa6d3%4010.0.27.106%3Bto-tag%3D1C951DC1-EB497A86%3Bfrom-tag%3Drp_5OC&REQUIRE=replaces&X-sipX-Authidentity=%3Csip%3A%7E%7Eid%7Epark%40inwsip.lcsnw.org%3Bsignature%3D50C7E6D8%253A%253A05936d0a35fdf946cde62d9a14e3e446%3E&REFERENCES=3bffb58-7f000001-13c4-2e2d5b-cf3f8dba-2e2d5b%4010.0.31.4%3Brel%3Drefer
> >
>
> The call-id is not seen anywhere else in this debug meaning, that portion
> of the call is not on the ADTRAN.  The ADTRAN accepts the Refer to it is up
> to the other device to tear that call down.
>
> Regards,
> Geoff
>
>
> -Bryan Anderson
>
>
>
> On Sat, Dec 8, 2012 at 12:19 AM, Ali Ardestani <[email protected]>wrote:
>
>>  You need to install ngrep and do an ngrep -Wbyline port 5060 to trace
>> sip messaging
>>
>>
>> On Friday, December 7, 2012, Bryan Anderson wrote:
>>
>>> So I pulled some tests and some traces, and have also opened a ticket
>>> with Adtran.  Using a Dialog Count I identified the following callid's to
>>> be in reference to one call.
>>>
>>> I called into the system with my cell  to an alias (1811) on the
>>> extension(5063).  I was parked on x6000. Extension 1802 retrieved the call
>>> and then we ended. (4=adtran 5=SipXecs, 181=x5603, 147=x1802)
>>>
>>>             61       INVITE   2012-12-07T20:05:51
>>> [email protected] => 4311 ->
>>> 1811(5063) -> 6000
>>>             30       INVITE   2012-12-07T20:06:21
>>> [email protected] => MoH
>>>             30       INVITE   2012-12-07T20:06:27
>>> [email protected] => 6000
>>>             33       INVITE   2012-12-07T20:06:28
>>> [email protected] => 6000
>>>             30       INVITE   2012-12-07T20:06:51
>>> [email protected] => 1802 -> 6000 retrieve call
>>>             29       INVITE   2012-12-07T20:06:52
>>> [email protected] => (From 1802
>>> retrieving the call) 4311 -> 1802
>>>
>>> Where does the BLF for the call park get so it doesn't release?  That is
>>> what I am looking for.  I am sorry I just don't know the SIP protocol well
>>> enough to find it on my own.  I have seen this with Polycom firmwares
>>> 3.2.4, 3.2.7 and yes 3.3.0 and 4.0.3(this trace).  I have a debug off the
>>> adtran to send to them, but they asked me were it is failing and I just
>>> don't know for sure myself.
>>>
>>> -Bryan Anderson
>>>
>>>
>>>
>>> On Fri, Dec 7, 2012 at 11:16 AM, Bryan Anderson <[email protected]>wrote:
>>>
>>> Thanks for the reply and I will defiantly test it.  We use a T1 for
>>> service into the Adtran and the Adtran is in SipXecs as an unmanaged
>>> gateway.
>>>
>>> -Bryan Anderson
>>>
>>>
>>>
>>> On Fri, Dec 7, 2012 at 11:07 AM, Ali Ardestani 
>>> <[email protected]>wrote:
>>>
>>> This is how we implemented call park with polycom and it works (it is a
>>> workaround though)
>>>
>>> 1. Extension 700 forwards to 701, 702, 703 and 703
>>>
>>> 2. added the below to the custom config of the phones (this is done so
>>> that the key does not timeout after 1 minute and call the park orbit
>>> directly
>>>         <call
>>>         call.offeringTimeOut="3600"
>>>         call.directedCallPickupMethod="legacy"
>>>         call.parkedCallRetrieveMethod="legacy"
>>>         >
>>>
>>> 3. subscribe to the presence of 700 on user speed dials
>>>
>>> 4. Make sure you use the bridge, we had problems with the call unparking
>>> when we did not use the bridge for incoming calls from trunk provider
>>>
>>> *5. Our firewal does ALG, so we had to uncheck "Use public address for
>>> call setup" under Devices=>Gateways(choose the gw)=>ITSP Account, This
>>> fixed our problem with the calls unparking, maybe your firewall is also
>>> doing some form of ALG*
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Fri, Dec 7, 2012 at 10:06 AM, Bryan Anderson <[email protected]>wrote:
>>>
>>> So I noticed some talk in a previous email "Call forward fails to
>>> external number" about the Adtran 900 series.  I have a couple of comments
>>> and questions.
>>>
>>> We have a TA908e 2nd gen running AOS A5.02.00.E.  We currently have not
>>> noticed any issue with having an external caller forwarded to and
>>> external number.  cell => user for 30sec => external number.
>>>
>>> What we have had issues with is presence monitoring of call parking.   We
>>> have a Polycom Soundpoint IP 650 with a sing side car that monitors 
>>> parklines 6000
>>> - 6003.  We can park calls no problem, and so far have not had trouble
>>> retrieving calls.  Our problem is that once the call gets retrieved
>>> from the call park the BLF never stops blinking.  I have to restart the
>>> Park/Presence servers.
>>>
>>> This is with SipXecs 4.4.0.
>>>
>>> Thoughts and comments would be appreciated.
>>>
>>>
>>> -Bryan Anderson
>>>
>>>
>>> _______________________________________________
>>> sipx-users mailing list
>>> [email protected]
>>> List Archive: http://list.sipfoundry.org/archive/sipx-users/
>>>
>>>
>>>
>>>
>>> --
>>> --
>>>  Ali S Ardestani
>>> Telephony Systems Engineer
>>> Private National Mortgage Acceptance Company (PennyMac)
>>> 6101 Condor Drive
>>> Moorpark, CA 93021
>>>
>>>
>>
>> --
>> --
>> Ali S Ardestani
>> Telephony Systems Engineer
>> Private National Mortgage Acceptance Company (PennyMac)
>> 6101 Condor Drive
>> Moorpark, CA 93021
>>
>> (805) 330-6004 Office
>> (818) 224-7442 x2654 Office
>> (626) 817-3512 Mobile
>> (818) 224-7397 Fax
>>
>> [email protected]
>>
>>
>> _______________________________________________
>> sipx-users mailing list
>> [email protected]
>> List Archive: http://list.sipfoundry.org/archive/sipx-users/
>>
>
>
> _______________________________________________
> sipx-users mailing list
> [email protected]
> List Archive: http://list.sipfoundry.org/archive/sipx-users/
>



-- 
Josh Patten
eZuce
Solutions Architect
O.978-296-1005 X2050
M.979-574-5699
http://www.ezuce.com
_______________________________________________
sipx-users mailing list
[email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-users/

Reply via email to